HomeIncident library
Incident library

INCIDENT REPLAY · 2024-02-22

AT&T Mobility nationwide wireless outage

At 2:42 AM an employee placed a new, misconfigured network element into production during a routine night maintenance window to expand capacity. Downstream elements propagated the error and three minutes later the network entered a protect mode that disconnected all devices from voice and 5G data. The rollback took close to two hours, but full restoration took at least 12 hours because device re-registration overwhelmed the registration systems. The FCC counted more than 125 million devices affected, more than 92 million voice calls blocked and more than 25,000 calls to 911 call centers prevented.

Required flowCapacity / signalling storm after changeCatalog confidence: high

A reconstruction, not the operator's network. Hammerhead's authors rebuilt the change from the public record as a small network of their own. No configuration here is the operator's. What follows is what the tool reported on that reconstruction; it is not a claim that the tool would have prevented the outage.

What happened

Who: AT&T Mobility (AT&T AS7018; FirstNet also affected)

The summary at the top is paraphrased from the sources below. Hammerhead's catalog rates its confidence in this account high.

Sources

What the public record says

  • At 2:42 AM an employee placed a new, misconfigured network element into production during a routine night maintenance window to expand capacity. Downstream elements propagated the error and three minutes later the network entered a protect mode that disconnected all devices from voice and 5G data. The rollback took close to two hours, but full restoration took at least 12 hours because device re-registration overwhelmed the registration systems. The FCC counted more than 125 million devices affected, more than 92 million voice calls blocked and more than 25,000 calls to 911 call centers prevented.
  • Who: AT&T Mobility (AT&T AS7018; FirstNet also affected)

What the reconstruction assumed

Decided by Hammerhead's authors to build a network that runs. None of it describes the operator's network.

Everything below is ours.

  • The family template (tools/replays/gen_family_policy.py): RAN-GW-1 (the radio access side) reaches CORE-1 (the mobile core) through NEW-ELEMENT-1 (the element inserted into the network without peer review), which filters inbound on its outside interface. The intent file names the flow: device registration reaches the core on udp/5060.
  • The FCC report says a configuration error on a newly inserted element, pushed without peer review, and does not name the element. Modelled as a new element in the signalling path whose filter, meant to block one unused port, blocks the registration signalling instead.
  • Cisco IOS syntax and stand-in address space throughout. A reconstruction, not any operator's configuration.

What Hammerhead said

Each run below is pinned by Hammerhead's replay tests at commit 0ee9732: the verdict and the quoted text are checked against the tool's output, and a run that disagrees fails the test. Reason lines are copied verbatim; a pinned line can be part of a longer reason. Every preflight run uses --max-blast-radius 100, so a change is never counted as caught merely because it touched many devices of a small network.

the change as intended

now: pass

On the reconstruction, hammerhead preflight on a before and an after snapshot with --policy @att_mobility_2024/policy.yaml --allow-flow-changes returned pass.

First run, before the fixes this batch prompted

first: pass

Pinned in tests/gauntlet/replays-policy.yaml.

the change as pushed

now: fail

On the reconstruction, hammerhead preflight on a before and an after snapshot with --policy @att_mobility_2024/policy.yaml --allow-flow-changes returned fail, for the reasons below.

Reason text pinned (some reason contains each line)

  • policy gate: high/critical failure(s) on the 'after' network: device registration reaches the core
  • NOT reachable (expected reachable)

First run, before the fixes this batch prompted

first: pass

Pinned in tests/gauntlet/replays-policy.yaml.

What changed in Hammerhead because of it

From the “What was fixed” table in Hammerhead's replay notes, where each fix landed after a failing test. “Found by” is copied as written; a family name means the batch this incident belongs to.

A flow rule names its protocol and port

Every reachability rule probed TCP/80, so an intent about SIP could not tell the push from the intended change: 7 of 11 intended changes failed on main and the one that dropped only the named port passed. protocol: and dst_port: on the rule; the policy gate's reason now carries the violation, not only the policy's name.

Found by: Required-flow family

Limits

  • The configurations are a reconstruction. Where the record is silent, the assumptions above fill the gap; if one is wrong, the replay is wrong in that respect.
  • A verdict describes the reconstructed network's converged state. Timing, traffic load, and anything a device does that no configuration states, such as software reacting to a health check or a firmware bug, are outside it.
  • A pass is reported as a pass. A fail says a gate fired on the reconstruction, nothing more.
  • Why the catalog counts this outage as partly in scope: The trigger was a configuration error on a newly inserted element pushed without peer review, which is the change-gating use case, but the element type is undisclosed and appears to be mobile-core rather than IP routing, and the 12-hour duration came from a device re-registration storm no config model predicts.

Imported from docs/replays/att-mobility-2024/FACTS.md and the replay manifests in the Hammerhead repository at commit 0ee9732.